Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

sensor to follow a person

Status
Not open for further replies.
Hi All

I am looking for advice on suitable components to use in a sensor circuit.Let me explain!!

You know in those comedy horror films where a portrait hanging on a wall suddenly moves it's eyes when a person walks by? well that's what I want to create.You may ask why? the answer is just a bit of fun!!

I was thinking of either using a small motor system to move the eyes or a solenoid system.

My question is really about the sensing and how best to do this.With my basic level I would probably go out and purchase a couple of cheap ready built IR.sensors and place these external to the portrait in positions that would make the eyes move when someone walks by.I do think there are better solutions to this and one that I could keep integral to the frame of the portrait making it one workable unit.

Could any one help with suggestions on this.

Cheers!!!!! wombweller
 
There is an article in the current Nuts & Volts Magazine on using sonar (ping sensors) to do a similar project
 
wombweller, see if you can follow this link Nuts and Volts - May 2009

It might be members only, but give it a try anyway.

Article starts on pg16


Hi Mike got to it ok, I think!! in the personal robotics section?cheers!!

I will have a read to see if it's any good for my application.

I was thinking along the lines of IR.emitter alongside a receiver. I thought if I had 3 sets 1 to the right 1 to the left and one one the middle maybe I could get the eyes to move as someone walked by.I'm not sure if these will do what I'm suggesting or if there are better alternatives??

I'm sort of hoping that someone will advise!!

Regards Wombweller
 
wombweller, see if you can follow this link Nuts and Volts - May 2009

It might be members only, but give it a try anyway.

Article starts on pg16

Hi Mike

Had a good look at those and they do seem to be very good.I will dig deeper on using ping sensors via google and see if I can find a useful design.

Can you think of alternative solutions? I was thinking of IR. transmitter and reciever but I think that you would have to be close for them to work.

Cheers!!
 
Last edited:
use a PIR sensor. PIR somes from "Passive InfraRed sensors" and is made for sensing humans :D. It sounds high tech, it may be, but a PIR can be found as low as 5 euro complete with it's control circuit ;)
 
moving eye

use a PIR sensor. PIR somes from "Passive InfraRed sensors" and is made for sensing humans :D. It sounds high tech, it may be, but a PIR can be found as low as 5 euro complete with it's control circuit ;)


Hi and thanks for the reply.

wonder if I could purchase already built PIR's and then remove the circuit from the case.My worry with that is the PIR won't sense productively without the lens.

The sensors have to be small and discrete and integral to the fame of the portrait.

The most important issue with this is the sensing needs to be in three stages- left,middle,right the idea being that when a person walks by each sensor picks them up as they pass by and the eyes following them as they do.

Creating a steady movement of the eyes may prove to be difficult and not
sure how to best achieve it so some advice here may be worthwhile.


Cheers!!
 
The success of this project depends on a couple of different things. If it is to be a stand alone unit, meaning there will not be sensors placed anywhere but on the "picture" it makes things a bit more difficult.
If you could place the sensors along side of the "picture" then IR beams could be used and the eyes would move in response to which beam was broken. This might take several sensors, otherwise the eyes might move in a chopy fashion.

PIR would detect movement but is generally a wide angle window of detection, so the eyes might not know which way to look.

Ping is only good for about 10" so you would have to be pretty close to get it to work.
 
Conceptually you could use two small stepper motors, one to move the eyes and one to scan the scene with an IR laser. You count the steps of the scanner stepper and when you get a return that indicates a person, you would then know how many steps the eyes need to move. This would give a fairly smooth tracking of the target.

The two steppers would need to be controlled by a micro-controller to do the necessary logic. You would need some type of limit switch on each stepper so you would know where the zero positions are.
 
Conceptually you could use two small stepper motors, one to move the eyes and one to scan the scene with an IR laser. You count the steps of the scanner stepper and when you get a return that indicates a person, you would then know how many steps the eyes need to move. This would give a fairly smooth tracking of the target.

The two steppers would need to be controlled by a micro-controller to do the necessary logic. You would need some type of limit switch on each stepper so you would know where the zero positions are.

Hi Carl

Sounds like an interesting solution!!

Are you saying that the IR.laser beam would be bounced back to a receiver? if not how would the signal be sent back to the PIC. ?
By using a PIR . sensor I can see how this would work as is already set up to do the signaling when activated. I do though like the sound of laser tracking as would be more precise.Would have to be an invisible Laser???

I get your point about using the stepper motors and can understand how with the right code the eyes would move smoothly.I'm guessing that both motors would have to track also as when a person is picked up the laser has t keep on it or the lase motor as to move fast enough to pick the person back up???

I figure we are now in the realms of simple robotics!! I will do some googling to see if I can find anything useful.If you can point any search queries that may help me along the way, also there may be some Laser tracking threads on this forum!! I will look.

Hopefully too I may find some coding as I do have a Programmer but wouldn't have a clue where to start with the code.

Lots of question, sorry!!

Cheers!!!

Mark
 


Hi

Cheers for the link.

I'm thinking that if I were to use PIR. sensors I would have to use a row of narrow detection ones or at least restrict their view.By limiting the angle of view and by using a row of them I just might get a smooth movement of the eyes.

The question is and one I'm looking at is will they still work of without the Fresnel lens!!

Cheers!!
 
Hi Carl
Are you saying that the IR.laser beam would be bounced back to a receiver? if not how would the signal be sent back to the PIC. ?
By using a PIR . sensor I can see how this would work as is already set up to do the signaling when activated. I do though like the sound of laser tracking as would be more precise.Would have to be an invisible Laser???

I get your point about using the stepper motors and can understand how with the right code the eyes would move smoothly.I'm guessing that both motors would have to track also as when a person is picked up the laser has t keep on it or the lase motor as to move fast enough to pick the person back up???
Yes, you would need a receiver sensor to detect the signal bounced back from the target. You could use a visible laser, but then you can have more problems with ambient light. (You might also be able to use an IR LED if you can focus it into a narrow enough beam).

The laser scan motor would continually scan (say a few scans per second) so it can keep track of the target motion. You just detect the position of the track motor when you detect the target and then put that same position into the eye motor. For every scan you would look at the difference between the scan target sense position and the previous eye motor position. If there is a difference you would move the eye motor accordingly. If you get no target signal then you would move the eye motor to a neutral position, such as looking straight ahead.
 
If you use the Sharp "GP2" distance sensing IR units, the ones with about a 2 yard range. Two of them facing diagonally inwards could fit on the sides of the picture frame and by testing the distance each sensor reports to the target you could get a pretty good X position of the person in front of the picture.

There are a few models of these sensors here;
(Pololu sensors) Pololu - Sensors

along with some ultrasonic sensor options and some tasty gyros/accelerometers etc etc. Well worth a look. :)
 
A few Haloweens ago, I bought exactly what you are looking for. Its portrait of a old hag in a frame with a pair of motorized eyes through cut outs in the picture. What it uses is pretty simple...and it doesnt really follow you. It just looks to the left when there are people to the left, and looks to the right when there are people to its right. The eyes would default to a center position when nothing was detected. It uses the same sensor that a lot of toys use, a small photocell at the end of a long black tube, connected to a simple 2 transistor circuit. 1 photocell is pointed in each direction, and senses abrupt changes in ambient light.
 
Last edited:
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top